IRS27951 / IRS27952

[nigelwright7557]

I got some more 27952 through the post today.

I decided not to solder one in but use a zif socket into an 8 pin DIP on my pcb.

First of all it wouldn't work at all. I eventually found my zif socket is slightly too narrow and the 27952 wasn't seating properly.

Then it powered up but gave out the wrong waveform. This turned out to be my bench power supply not supplying enough current to run normally. This means I might have been throwing out perfectly good 27952's when they didn't power up properly after soldering in.

I think I will now wait until I get better pcb's through with larger pads on the 27952, this will allow pushing solder along the pad to the pin instead of heating up the 27952 at the pin. I suspect my soldering iron is far too hot for this. I have a large tip for soldering larger through hole components. I will see if I can find a smaller tip.

[rikkitikkitavi]

I think you have a good plan. First get a good PCB then a better soldering iron. I have one with a 0,8mm tip with temp control and of the 50+ SO chips none has failed sofar due to heating atleast.I solder pin by pin with fine.solder wire (high Q) no wick to remove excess solder and use extra liquid fluss , after I wash with isopropylalcohol.

Using a ZIF socket with these circuits doesnt sound right, will parasitic capacitance not affect.drive capacity etc, and general function.

But I encourage your courage to continue. Every morning I check this thread to follow your progress in the right direction!! Go go go!

thanks for sharing

[nigelwright7557]

It will take a couple of weeks for the new pcb's to come through.

Luckily I am always up for a challenge in electronics/software.
There have been a lot of mistakes but I am learning from them which was the object of the exercise.

You would probably laugh if I told you I had been in electronics for 30+ years.
Many times I was given things to do I hadn't done before but mostly came through with reasonable solutions in the end.

I have also bought an isolation transformer off ebay so I can look at the primary side and see what is going on. I have been working half blind up to now only seeing what was happening on the secondary side.

[hilli_billi]

Hi Nigel,
I have my oscilloscope connected via an isolation transformer. The patient runs off mains.
I don't like that, when measuring - it's so dangerous.
When I had problems with the regulation loop, I wound me a special transformer with 10 times secondary and supplied the mains input of the SMPS from a Lab-supply. So I can safely measure.
Later I switched back to the 'right' transformer of course.

[nigelwright7557]

I run mine off 30 volts dc to start with and apply the 30 volts through a 390r resistor to the 27952. That way I can check for anything seriously wrong. If the 27952 starts up with a square then it is usually ok to apply full mains.

[nigelwright7557]

I decided to have a quick go at soldering in a 27952.
I soldered it in but got a blob of solder across two pins, so I wicked it off as quick as I could. I buzzed out the ic to look for shorts or open circuits and it was fine.
I powered it up off 30 volts but it was totally dead.

I had a dig around in my junk room and found an old temperature controlled soldering station. I set the iron to be just hot enough to melt solder (about 3/4 heat setting). I soldered in another 27952 and got a blob across two pins again. I wicked off the blob. This time the 27952 powered up fine.

It looks like my usual soldering iron is simply far too hot for smd work.

I powered up my pcb off mains and the output looks fine.

[rikkitikkitavi]

It is not the heat but the time , because when the.solder starts to melt temperature stays constant until.all is melted, after which temperature.increases again. I have my soldering iron set to 390 C (leaded tin) and a tiny chisel tip.
of course if the ironhas high power and a largish tip the temperature plateu is reached and passed over quickly and temperature rise fast. hard.tell.
all I can saybis that I am very very happy with my Ersa Micro60 solder station.

Good luck to you, can we see some pics perhaps?

[nigelwright7557]

My previous soldering iron had a large tip because I had been soldering inductors on a class d amplifier.

[nigelwright7557]

The wire is a fix to a ground loop, not sure now if it is still needed but my new pcb sorts that out.
Still a couple of components sticking up for slight changes and testing.

[nigelwright7557]

For some reason the 27952 blew up tonight during testing. I was taking quite a few amps when it stopped working.
Soldered in a new chip and of course that didnt work either.
So using liquid flux I soldered in another 27952 and that worked fine.
The flux seems to stop the solder blobbing so much and seemed to make things solder quicker.